HOW TO INTRODUCE CREDIT SCORING FOR...
Transcript of HOW TO INTRODUCE CREDIT SCORING FOR...
www.lisim.biz
MFI’s CHALLENGES TO USE CREDIT SCORING
Challenge•Change Resistance
•Lack of Development Vision
•Scoring as a Threat
Cultural Constraint
Databases Availability
Solution•Scoring as a tool for
challenging goals
•Intensive Training for all related staff
Challenge•Not enough quantity and
quality data bases
Solution•Expert models
www.lisim.biz
MFI’s CHALLENGES TO USE CREDIT SCORING
Challenge
•Not enough and specific technological resources for the
microfinance market
Technological Resources
Solution
•Technological tools may be adjusted from other massive
segments
Key factor for a profitable portfolio at microfinance
www.lisim.biz
HOW DOES CREDIT SCORING AFFECT POLICIESPolicies Review
Scoring helps to validate aspects inside policies and improve all the approval process.
Necessity
ValidationHigh Risk
Low Risk
Statistical Evaluation
AvoidTo accept risky population andto miss target population
Policies Target Population
Decision Real Risk Profile Policy EvaluationLow Risk Wrong PolicyHigh Risk Right PolicyLow RiskHigh Risk
Total 100% Average Risk
Policies Population Target
Rejected Population
Approved PopulationRight PolicyWrong Policy
www.lisim.biz
HOW DOES CREDIT SCORING AFFECT POLICIES
A
R
R
R R
R
R
A = Approve
R = Reject
High Risk = Rejection
Average / High = Disbursement with short term, low amount, high interest.
Average / Low = Disbursement with average/short term, average/low amount, average interest rate
Low Risk = No restrictions.
Looks Right Clients
One single characteristic at
a time
Multiples characteristicsPolicies Scoring
Policies ScoringTo Current Products
Looks Right Products
To Potential Clients
Total # Range score # PERS % PERS
A 4000 43.48% 9200 2.47%B 2500 36.76% 6800 1.82%C 5800 33.72% 17200 4.61%D 10100 27.01% 37400 10.03%E 9100 24.80% 36700 9.84%F 5200 13.33% 39000 10.46%G 4500 10.47% 43000 11.53%H 2700 5.74% 47000 12.60%I 1700 4.15% 41000 10.99%J 770 1.73% 44500 11.93%K 450 0.88% 51200 13.73%
46820 12.55% 373000 100.00%
AVERAGE LOW RISK
LOW RISK
HIGH RISK
TOTAL
Bad Total %
AVERAGE HIGH RISK
*Scoring keeps only basic policies
www.lisim.biz
Start
First data collection (enough as to apply
basic policies)
Basic and Bureau Policies application
Does the client fits the policies
Sociodemographic and Business Data
Collection
1
Application Score
Calculus
Yes
NoBehavior Database
3
1
High Risk (Automatically
Rejection)3Yes
No
Average Risk?
Yes
Intensive Verification (Second Visit +
Reference consulting)
Light Verification (Reference consulting)
No
Match Verification?
2
Yes
2
Notification and Disbursement
End
Clients Database3
3No
HOW DOES CREDIT SCORING AFFECT POLICIES
Approval Process (Scoring Evaluation methodology)
Scoring
Segmented RiskEvaluation
Traditional Process
Standard Evaluation
www.lisim.biz
CLIENTS RELATIONSHIP TROUGH SCORING
CLIENT CYCLE
Approval Process:Applications Models for new clients; “More and best new
clients”
Collection Process:
Behavior models to segment population and apply
differentiated strategies regarding payment probability.
Renewal and Cross SellsProcess: Behavior Models to
improve the renewal and cross sells process regarding risk and
consumption tendencies
Provisions estimation:Behavior models for an accurate
expected loss estimation. (as included in BASEL II)
www.lisim.biz
SCORING SISTEMS MAINTENANCE
Implementation Stage It’s extremely important to validate integrally the right implementation
model.
Monitoring Reports•Behavior Stability
•Scorecard Evaluation
•Data Quality
•Rejection Levels
•Scoring Impact
Maintenance Process• An integral review of the current variables
behavior
• An evaluation of additional variables for possible inclusion in the model.
A maintenance is required depending on an evaluation reportalert
www.lisim.biz
CONCLUSIONS
•Scoring implementation challenges have always alternatives of success.
•Scoring gets more attention to clients, due to it maximizes the approvals by reducing the policies and customizing the services offered to the client’s risk.
•Integral risk management through predictive CRM.
•The maintenance is important to guarantee the appropriate performance of the Score Card.
www.lisim.biz
THANK YOU VERY MUCH FOR YOUR ATTENTION
For additional information, please do not hesitate to contact me:
Lilian SimbaquebaLiSim C.E.O
[email protected]: 57 1 602 20 20
Address: Cll. 94 A No. 11 A – 39Bogotá - Colombia
www.lisim.biz
OUR COMPANY
LiSim COLOMBIACalle 94 A No. 11A-39
Tel : 57(1) 6022020 Bogotá D.C. Colombia
LiSim MEXICO S. De R.L. de C.V.
San Borja, 816-I. Colonia del Valle.C.P. 03100 Teléfonos: (52 55) 5559-7437
México, D.F.
E-Mail: [email protected]
LiSim ECUADORAv. La República 307 y Diego de Almagro Oficina 7B. Tel: 593-2-2529341/2550459 Quito
LiSim PERUAv. José Pardo 223, Piso 9, Miraflores.
Tel: 511 2160016 Fax: 511 3722518Lima
LiSim BOLIVIACalle Benito Juárez #48 Casa 5 La FloridaTel (5912) 274 0753La Paz
LiSim CHILESimón Bolívar 5917 M Teléfono: (56 2) 2278118Santiago de Chile
LiSim ARGENTINAReconquista 661 – Piso 4 C1003ABMTel (5411) 43156700Buenos Aires